The ability to non-invasively predict which cancer patients will respond to chemotherapy will have significant implications for patient care and public health with the potential to spare non-responding patients the high morbidity and cost associated with these treatments. Therefore, novel non-invasive imaging technologies to allow both prediction and/or early evaluation of treatment response are critically needed. Due to its wide accessibility, lack of radiation, relatively low cost, and excellent visualization of the live, ultrasound perfusion imaging following intravenous administration of contrast microbubbles is a promising new functional modality for assessing tissue perfusion of liver metastases, the most common site of metastases in patients with colorectal cancer. However, current two-dimensional dynamic contrast-enhanced ultrasound (DCE-US), covering only a small fraction of the tumor volume, is fundamentally limited in quantification by the three-dimensional (3D) heterogeneity of tumor perfusion, resulting in consecutive sampling errors with substantial over- or underestimation of treatment response on DCE-US imaging. These limitations could be overcome by a novel 3D DCE-US imaging approach combining innovative matrix array transducer technology with real-time electromagnetic tracking to provide motion-compensated, accurate quantitative and volumetric assessment of tissue perfusion of liver metastases. Our overall goal is to assess 3D DCE-US reproducibility in patients with colorectal liver metastases and to evaluate its ability to predict treatment response. In 50 patients with newly diagnosed liver metastases from low grade colorectal adenocarcinoma, the reproducibility of tracking-assisted 3D DCE-US will be assessed by using a disruption-replenishment DCE-US technique which we had validated in previous preclinical experiments in human colon cancer xenografts in mice. Baseline and early changes of tumor perfusion parameters (obtained before and at 2-3 weeks after the initiation of chemotherapy) will then be correlated to the following endpoints: local treatment response, progression free survival, and histology. Treatment response as defined on abdominal CT scans acquired at 2 months following treatment initiation will be determined using RECIST 1.1 reporting as reference standard. Baseline values and early changes in perfusion parameters will be correlated with CT findings and with histology in patients undergoing surgical liver metastasis resection.